1895.] The Sparlc Spectrum of Argon as it appears in Air. 293 



IV. " On the Spark Spectrum of Argon as it appears in the 

 Spark Spectrum of Air." By W. N. HARTLEY, F.R.S., 

 Professor of Chemistry, Royal College of Science, Dublin. 

 Received January 31, 1895. [Revised February 18, 1895.] 



The spark spectrum of air as photographed, mapped, and described 

 by Hartley and Adeney* contains various lines which they have not 

 been able to attribute to oxygen or to nitrogen, having no grounds 

 for assigning them to one element or the other. The lines belonging 

 to oxygen and nitrogen, when produced by uncondensed sparks, are 

 comparatively well known ; so also are the lines of oxygen when a 

 condensed spark is used, but it is otherwise with the lines of nitrogen 

 when the spark is condensed. 



It was observed by one of usf that the very abundant air lines in 

 the photographed spectrum of air were subject to considerable varia- 

 tions according to circumstances. Thus, by modifications in manipu- 

 lation, the air- spectrum might almost be suppressed, though the 

 metallic lines remained strong. 



On the other hand, it was remarked that certain metallic electrodes 

 gave spectra of air with some lines more prominent than others, also 

 more sharply and distinctly defined. Aluminium and copper were 

 two which gave the most distinctly-marked air-lines. Platinum also 

 in some cases. 



From statements contained in the paper contributed by Mr. 

 Crookes on the spectrum of argon it may be gathered that the 

 following conditions have been observed to yield the brightest and 

 purest spectrum of argon : 



1. As to pressure, about 3 mm. 



2. Electrodes of aluminium or platinum. 



3. A condensed spark from a coil and Leyden jar. 



These conditions were very nearly approached when Hartley and 

 Adeney's spectra of air were photographed, thus: 



1. As argon is about 1 per cent, of the atmosphere, its partial 



pressure is about 7'5 mm. 



2. Electrodes of aluminium, copper, platinum, and palladium were 



used. 



3. A condensed spark was always employed, this being produced 



by using a coil and Leyden jar. 



It was therefore reasonable to expect that the spectrum of the 



* ' Phil. Trans.,' vol. 175, p. 63, 1884. 



f ' Journ. Chem. Soc.,' vol. 41, p. 84, 1882 j also ' Sci. Trans. Koyal Dublin Soc.,' 

 vol. 1, p. 231 (new series), 1881. 



